The article compares different methods for predicting Korean interest rates, economic growth, and inflation rates. They tested 94 models, including standard autoregressive and vector autoregressive models. The best forecasts for interest rates came from Bayesian vector autoregressions using certain variables. For growth rates, average forecast models were generally better, while autoregressive models were more accurate than vector autoregressive models. And for inflation rates, models with a recursive window performed the best.